Oral Candidiasis in the Precision-Diagnostics Era: A Critical Review of Conventional and Molecular Diagnosis, Antifungal Susceptibility, and Therapeutic Decision-Making

Abstract

Background: Oral candidiasis is usually diagnosed clinically, yet the same Candida species can be recovered from healthy mouths. Precision diagnosis must therefore distinguish carriage from active mucosal disease, select the correct anatomical sample, identify species only when clinically consequential, and interpret antifungal susceptibility in the context of exposure, biofilm, and host factors.

Methods: A structured critical review was conducted through 22 July 2026. PubMed-indexed primary studies, oral diagnostic-accuracy studies, randomized therapeutic trials, and authoritative WHO, NIH, IDSA, CLSI, and EUCAST documents were prioritized. Searches combined oral or oropharyngeal candidiasis with microscopy, culture, fungal burden, MALDI-TOF MS, PCR, sequencing, metabolomics, biosensor, susceptibility, MIC, resistance, recurrence, and therapy. Seminal studies before 2000 were retained when they defined treatment efficacy. Because case definitions, specimens, reference standards, and endpoints were heterogeneous, numerical results are presented as study-specific anchors rather than pooled estimates.

Results: Lesion phenotype and risk context remain the diagnostic gate. Direct microscopy adds evidence of fungal activity; culture enables recovery, mixed-species recognition, identification, and susceptibility testing but is not proof of infection. In 124 suspected cases, fluorescent staining outperformed 10% KOH for sensitivity (85.48% vs 64.52%), specificity (91.94% vs 72.58%), and AUC (0.887 vs 0.685). Concentrated oral rinse increased Candida recovery relative to swabbing, but high burdens can still occur in asymptomatic carriers. MALDI-TOF accurately identifies cultured isolates when databases and extraction are adequate; qPCR increases analytical sensitivity but may detect non-viable DNA and clinically silent carriage. Oral microbiome, metabolome, microfluidic, and paper-based biosensor studies are promising but remain discovery or prototype evidence. Standardized CLSI or EUCAST broth microdilution is appropriate for selected refractory, recurrent, high-risk, or non-albicans infections. Current EUCAST breakpoints are method- and species-specific; caspofungin lacks direct EUCAST breakpoints, and neither CLSI nor EUCAST provides validated nystatin clinical breakpoints. Oral biofilm tolerance and mucosal drug exposure are not captured by planktonic MIC alone. Randomized trials support fluconazole for moderate-to-severe disease and convenient topical regimens for selected localized disease, but recurrence depends strongly on immune state and unresolved reservoirs.

Conclusions: Precision oral mycology should be organized as a layered decision system: phenotype, lesion-linked fungal evidence, species identification, selective susceptibility testing, and therapy matched to severity, host state, prior exposure, and local reservoirs. The next generation of diagnostics must measure pathogenic activity—not simply fungal presence—and demonstrate incremental clinical utility beyond optimized examination, microscopy, and culture.